AI Becomes Standard in Banking, Say Google Cloud, R/GA

At the launch of the first Hispanic American edition of Google Cloud FinFacts 2026, Google Cloud and R/GA delivered a clear message to financial executives: artificial intelligence is no longer a differentiator in banking—it has become the new baseline.

Presented in Mexico City, the study evaluates 20 financial institutions across Mexico, Colombia, Chile and Argentina, analyzing more than 28 usability, infrastructure and service criteria within digital onboarding and product-contracting journeys. The findings reveal a sector undergoing active transformation but still far from meeting the expectations of AI-enabled consumers.

“We no longer have just a static snapshot of the market,” said Herbert Freitas, creative director, R/GA and one of the leaders behind the study. “In Brazil, where we have conducted this research for five consecutive years, we now see a historical evolution. And in the last two years, the changes have been especially profound because of artificial intelligence.”

The AI Expectation Gap

Consumers are already integrating AI into their daily lives, using generative tools to analyze expenses, summarize documents and guide financial decisions. Banking, however, has not yet fully caught up.

Customers, for example, are increasingly requesting a clearly visible button to export bank statements. “Because they want to upload that data into an AI tool to obtain deeper financial analysis. They are effectively building their own digital financial advisor outside the bank,” Freitas said.

Despite this behavioral shift, none of the 20 institutions evaluated offers conversational account opening. Onboarding remains largely a rigid, step-by-step process. Only two institutions provide mobile-optimized summaries of terms and conditions. Nine of the 20 interrupt the digital journey by redirecting users to external channels during account opening.

The study found that only seven institutions offer any form of interface personalization, and just two search engines can handle typos or semantic queries. Voice search functionality was absent across all evaluated platforms.

Regulation Is Not the Obstacle

One recurring question during the event was whether Mexico’s regulatory complexity and relatively low bancarization levels were slowing AI innovation in financial services.

David Ruiz, enterprise director, Google Cloud, pushed back on that narrative. “I do not see regulation as an obstacle,” Ruiz said. “I see it as supervision. Banks evaluate use cases, ensure quality, present them to regulators, and then move to production. AI adoption is already happening.”

According to the study, more than half of the institutions analyzed already deploy AI-powered chatbots, and five use AI to enable more fluid conversations. However, maturity remains uneven, as only four chatbots were able to adequately answer questions about contracted product terms, and none could resume a conversation after a five-minute pause.

Security, Ruiz emphasized, remains foundational. “Customer data belongs exclusively to our clients. No one at Google uses that data to train models, nor is it used for any other purpose,” he said.

He also highlighted the growing use of AI to counter AI-driven threats, noting that anomaly detection, behavioral fraud analysis and biometric verification enhancements are becoming standard components of cloud-based security architecture.

The Real Bottleneck: Data Silos

If AI is the engine, data integration is the fuel, experts say. A major challenge for implementing the technology lies in legacy infrastructure and fragmented data environments.

Many institutions still operate with siloed systems—core banking platforms running on mainframes, separate digital layers and disconnected product databases. This fragmentation results in redundant data requests, repetitive onboarding steps and limited personalization.

According to the study, 18 of the 20 institutions require customers to re-enter information when applying for a credit card, even if they already hold an account. “It feels strange to be treated like a stranger when they already know you,” Freitas observed.

Cloud infrastructure, Ruiz argued, is central to solving this challenge. A unified customer data platform enables what the industry refers to as a “single customer view,” reducing friction while strengthening risk management.